Professional Documents
Culture Documents
net/publication/300937828
CITATIONS READS
0 132
1 author:
Esha Raffie
Sri Krishna College of Arts and Science
32 PUBLICATIONS 27 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Esha Raffie on 11 April 2016.
Basic Assumptions
[10]
The quality characteristic is represented by a random Based on the OC function of the QSS Romboski (1969)
variable X measured on a continuous scale. the OC function of QSVSS (nTσ, nNσ; k) can be written as
Distribution of X is normal with mean and Standard
Deviation. (w T )
Pa (p) (3)
An upper limit U, has been specified and a product is 1 (w N ) (w T )
qualified as defective when X>U. [When the lower limit With
L is Specified, the product is a defective one if X<L].
The Purpose of inspection is to control the fraction
w T n Tσ (U k σ μ)/σ (v k σ ) n Tσ
defective, p in the lot inspected. w N n Nσ (U k σ μ)/σ (v k σ ) n Nσ
When the conditions listed above are satisfied the fraction v (U μ)/σ
defective in a lot will be defined by
Under the assumption of normal approximation of the non-
p= 1-F (v) =F (-v) central t distribution (Abramowitz and Stegun, 1964) [1], the
values of PN and PT are respectively given by
with v= (U-) /
PN F(w N ) pr[(U x)/σ k σ ] (4)
y 1
e Z /2dz PT F(w T ) pr[(U x)/σ k σ ] (5)
2
and F(y) (1)
2π Where, PN and PT are the proportions of lots expected to be
Where Z ~ N (0, 1). It is to be recalled here that the criterion accepted using normal (nNσ, k) and tightened (nTσ, k)
for the -method variable plan is to accept the lot if x + k variables sampling plans respectively. These two equations
σ U, where U is the upper specification limit or x - k σ are applied to the OC function of QSVSS (nTσ, nNσ; k). We
L, where L is the lower specification limit. The operating get the following
procedure of QSVSS (nTσ,nNσ;k) is described below.
Pr[(U x)/σ kσ ]
Pa (p) (6)
Operating Procedure 1 Pr[(U x)/σ kσ ] Pr[(U x)/σ kσ ]
The steps involved in this procedure are as follows If Pa(p1) and Pa(p2) are
Step 1. Under normal inspection, draw a sample of size nNσ
from the lot, inspect and record the measurement of Pr[(U x)/σ kσ ] (7)
Pa (p1) 1 α
the quality characteristic for each unit of the sample. 1 Pr[(U x)/σ kσ ] Pr[(U x)/σ kσ ]
Compute the sample mean x . Where x N = x i
nN Pr[(U x)/σ kσ ] (8)
Pa (p2 ) β
Step 2. i) if x N k σ U or x N - k σ L accept the lot 1 Pr[(U x)/σ kσ ] Pr[(U x)/σ kσ ]
and repeat step 1 for the next lot. The expression for the sum of Producer’s and consumer’s
ii) if x N k σ U or x N - k σ L reject the lot
risks is given by
go to step 3.
Step 3. Under tightened inspection, draw a sample of size nTσ α +β =1-Pa (p1) +Pa (p2 ) (9)
from the next lot, inspect and record the
measurement of the quality characteristic for each For given AQL and LQL, the parametric values of QSVSS
unit of the sample. Compute the sample mean x . namely k, the sample sizes nNσ, nTσ, α and β are determined
Where x T = x i . by using a computer search.
nT
Step4. i) if x T k σ U or x T - k σ L accept the lot A procedure for QSVSS (nTσ, nNσ; k) System involving
and repeat step 1 for the next lot. Minimum sum of the risks
ii) if x T k σ U or x T - k σ L reject the lot go From table 1, a procedure for designing of Quick Switching
Variables Sampling Systems involving Minimum sum of
to step 3. risks for the given values of AQL and LQL is indicated
Where nNσ and nTσ are the sample sizes of normal and below.
tightened sampling, variable plans respectively, and k is the Table 1 is used to select the QSVSS for given values of
acceptance constant under σ-method. Where x and σ are the (AQL, 1-α), (LQL, β). The plan given here have the
average quality characteristic and standard deviation minimum sum of the risks. Fixes the values of p1 and p2 from
respectively. The quick switching variables sampling system which a Quick Switching Variables Sampling System can be
as defined QSVSS (nTσ, nNσ; k). As thetightened plan sample
selected under known -method. Entering the row, giving p1
size nTσ is greater than the normal plan sample size nNσ, for and p2, one gets the acceptance criteria k, α+β and the sample
designing the QSVS system nTσ is fixed as a multiple of nNσ. sizes nNσ and nTσ of QSVSS (nTσ, nNσ; k). For example, for
i.e. nTσ = m nNσ, where m >1.
given p1=0.002, p2=0.007, nT=320, nN=160, k=2.852,
α=5% and β=0%.
Operating Characteristic Function
According to Romboski (1969) [10], the OC function of QSS
QSVSS with Unknown variables plan as the reference
is given by
plan
PT If the population standard deviation is unknown, then it is
Pa (p) (2)
1 P P
N T estimated from the sample standard deviation s (n-1 as the
~ 105 ~
International Journal of Applied Research
S (x i x) / n N 1
2
Fig 1: Normal, Composite and Tightened OC Curves with minimum sum of risks for QSVSS, (nT=352, nN=176,
k=2.733, α=9% and β=0%)
~ 106 ~
International Journal of Applied Research
Construction of Table values of nTσ, nNσ, kσ, α andβ areconstructed. A procedure for
The OC function of QSVSS (nTσ, nNσ; k) is given by finding the parameters of unknown σ - method plan from
equation (2). For specified p1 and p2 the equation (6) would known σ-method plan with parameters (nS, kTS, kNs), are
result in obtained by using computer search routine through C++
programme. The values of nT, nN, k for the QSVSS that
PT satisfying equation (2) are obtained using (10). By definition,
Pa (p) (10)
minimizing the sum of risks are determined as
1 PN PT
where α β 1 Pa (p1 ) (Pa (p2 )
wT 1 2
PT e z /2 dz (11) where,
2π
P (p ) Pr[(U x)/σ kσ ]
wN 1 a 1 α
2 1 Pr[(U x)/σ kσ ] Pr[(U x)/σ kσ ]
e z /2 dz
1
and P (12)
N 2π
Pr[(U x)/σ kσ ]
Using iterative procedure equations, (11) and (12) is solved Pa (p2 ) β
for given values of p1, p2, m, α and β to get the values of nNσ 1 Pr[(U x)/σ kσ ] Pr[(U x)/σ kσ ]
and kσ for the specified pair of points, say (p1, α) and (p2, β)
on the OC curve. Here, the values of m canbe taken m >1and The sample size nTσ equals mnNσ and nTs equals mnNs, and
find the desired parameters. In Table 1, provided, if m=2, the hence only nTσ, nNσ, k, α and β are tabulated.
Table 1: The values of nTσ, nNσ, kσ, nTs, nNs, ks, α, β for given AQL and LQL Involving Minimum Sum of Risks.
p1 p2 nTσ nNσ kσ α β nTS nNS kS
0.002 504 252 3.080 9 1 2895 1447 3.081
0.003 350 175 3.072 8 0 2001 1001 3.072
0.004 338 169 3.062 7 0 1922 961 3.062
0.005 326 163 3.072 6 0 1864 932 3.072
0.006 324 162 3.062 5 0 1843 921 3.062
0.007 314 157 3.062 4 0 1786 893 3.063
0.008 312 156 3.062 3 0 1774 887 3.063
0.009 310 155 3.062 2 0 1763 882 3.063
0.010 300 150 3.052 1 0 1697 848 3.053
0.020 198 99 3.012 1 0 1096 548 3.013
0.030 158 79 2.992 1 0 865 433 2.994
0.001
0.040 134 67 2.982 1 0 730 365 2.984
0.050 132 66 2.972 1 0 715 357 2.974
0.060 130 65 2.972 1 0 704 352 2.974
0.070 128 64 2.972 1 0 693 347 2.974
0.080 124 62 2.972 1 0 672 336 2.974
0.090 118 59 2.962 1 0 636 318 2.964
0.100 112 56 2.962 1 0 603 302 2.964
0.110 106 53 2.952 1 0 568 284 2.955
0.120 102 51 2.942 1 0 543 272 2.945
0.130 98 49 2.942 1 0 522 261 2.945
0.140 94 47 2.932 1 0 498 249 2.935
0.150 90 45 2.932 1 0 477 238 2.935
0.003 356 178 2.862 9 0 1814 907 2.863
0.004 344 172 2.862 8 0 1753 876 2.863
0.005 332 166 2.852 7 0 1682 841 2.853
0.006 330 165 2.852 6 0 1672 836 2.853
0.007 320 160 2.852 5 0 1622 811 2.853
0.008 318 159 2.842 4 0 1602 801 2.843
0.009 316 158 2.842 3 0 1592 796 2.843
0.010 306 153 2.832 2 0 1533 767 2.833
0.020 204 102 2.792 1 0 999 500 2.794
0.002 0.04 140 70 2.762 1 0 674 337 2.764
0.05 138 69 2.772 1 0 668 334 2.774
0.06 136 68 2.752 1 0 651 326 2.754
0.07 134 67 2.762 1 0 645 323 2.765
0.08 130 65 2.762 1 0 626 313 2.765
0.09 124 62 2.742 1 0 590 295 2.745
0.1 118 59 2.752 1 0 565 282 2.755
0.11 112 56 2.732 1 0 530 265 2.735
0.04 140 70 2.762 1 0 674 337 2.764
0.05 138 69 2.772 1 0 668 334 2.774
~ 107 ~
International Journal of Applied Research
~ 111 ~